I would say it depends on the burden of the disease in question. If hospitals are registering more cases than previously then it is an outbreak since the definition has shifted from endemic to epidemic.
The terminology of endemic and epidemic is mostly referred to infectious diseases which are known for quick spread like viral and bacterial infections. Therefore, the epidemic which also called an outbreak occurs in a short period of time which could vary from hours to few days depending of the type of disease. The example of that is the number of reported cases of diarrhea in children; if the regular reported cases on dialy or weekly bases has increased dramatically that will signal an outbreak of diarrheal disease which should be reported to the public health authority to investigate the cause and intervene.
Additionally a ln epidemic can occur when a disease emerged in a place it never was. Eg the re-emergence of a single case of small pox could be described as an epidemic
As far as period of surveillance is concerned, one must at least have last year data (collected from different hospital through passive surveillance) then compare it with the recent data, that is no of new cases. And then compare it as suggested by Dr Mamoona.
Thanks all for your responses.But Still The question is there .The defintion of outbreak is clear for me : it is the occurrence of cases of disease in excess of what would normally be expected in a defined community, geographical area or season. outbreaks of single cases is also clear for me . What is vague for me is as follows... eg : if the regular vaiation for a disease is about 100 -200 per year when to say that it is an outbreak ?? if it exceeds 250 ?? or 300 or 400 ?? how to confirm that this is an outbreak ?? at what level ?? and on what basis ?? is there any reference or general rule ???
To answer your question above, I think the thresholds for different diseases are determined by various local and international health authorities. E.g. if there is one smallpox case it would be an outbreak since it has been eradicated. If there are 2 or 3 more unrelated food poisoning cases compared with last year, you might not consider it an an outbreak. Then again, if you see a cluster of food poisoning cases in a specific area, you might consider that an outbreak. So to decide whether an increase of cases is an outbreak, we might consider the transmissibility of the pathogen, the R0, the severity of the disease, and whether this is an artifact (or increased testing etc). To this end, this paper may be interesting to read: http://journals.plos.org/plospathogens/article?id=10.1371/journal.ppat.1003277